I. Acoustic study: Methodology
5. Data labeling and analysis

Formant values were normalized using Watt &
Fabricius S-centroid procedure (Watt &
Fabricius 2002).
1
I. Acoustic study: Methodology
5.1. Raw data
2600
2100
F2
1600
1100
600
200
300
Vocal i Herrero (2010)
Vocal e Herrero (2010)
400
Vocal E Herrero (2010)
500
Vocal a Herrero (2010)
600
700
Vocal O Herrero (2010)
800
Vocal o Herrero (2010)
900
Vocal u Herrero (2010)
1000
2
I. Acoustic study: Methodology
5.1. Raw data
2600
2100
F2
1600
1100
600
200
Vocales del hablante 1
300
Vocal i Herrero (2010)
400
Vocal e Herrero (2010)
500
Vocal E Herrero (2010)
600
700
Vocal a Herrero (2010)
Vocal O Herrero (2010)
800
Vocal o Herrero (2010)
900
Vocal u Herrero (2010)
1000
3
I. Acoustic study: Methodology
5.1. Raw data
2600
2100
F2
1600
1100
600
200
300
Vocales del hablante 1
Vocales del hablante 2
Vocal i Herrero (2010)
400
Vocal e Herrero (2010)
500
Vocal E Herrero (2010)
600
Vocal a Herrero (2010)
700
Vocal O Herrero (2010)
800
Vocal o Herrero (2010)
900
Vocal u Herrero (2010)
1000
4
I. Acoustic study: Methodology
5.1. Raw data
2600
2100
F2
1600
1100
600
200
Valores medios hablantes 1 y
2
Vocales del hablante 1
300
Vocales del hablante 2
400
Vocal i Herrero (2010)
500
Vocal e Herrero (2010)
600
Vocal E Herrero (2010)
700
Vocal a Herrero (2010)
Vocal O Herrero (2010)
800
Vocal o Herrero (2010)
900
Vocal u Herrero (2010)
1000
5
I. Acoustic study: Methodology
5.1. Raw data
2600
2100
F2
1600
1100
600
200
300
400
Valores medios hablantes 1 y
2
Vocal i Herrero (2010)
Vocal e Herrero (2010)
500
Vocal E Herrero (2010)
600
Vocal a Herrero (2010)
700
Vocal O Herrero (2010)
800
900
Vocal o Herrero (2010)
Vocal u Herrero (2010)
6
1000
I. Acoustic study: Methodology
5.2. Data normalization
For each speaker:
F1, F2 [i]
F1, F1 [i]
S (F1, F2)
F1, F2 [a]
7
I. Acoustic study: Methodology
5.2. Data normalization
For each speaker:
Normalized values:



Normalized_F1_[e] = Observed_F1_[e] / S1
Normalized_F2_[e] = Observed_F2_[e] / S2
Intra-speaker variation is in both
sides of the division; hence, with
normalized values, its effects
are eliminated.
8
I. Acoustic study: Methodology
5.2. Data normalization
Ex. from one particular speaker:
[i] = (0,75, 1,67)
[u] = (0,76, 0,59)
S (1, 1)
[a] = (1,49, 1,08)
9
I. Acoustic study: Methodology
5.2. Data normalization
F2
1.90
1.40
0.90
0.40
0.40
0.60
Hablante 1, Norm
0.80
1.00
1.20
1.40
Hablante 2, Norm
1.60
1.80
10
2.00
I. Acoustic study: Methodology
5.2. Data normalization
F2
1.90
1.40
0.90
0.40
0.40
0.60
Media hablantes 1 y 2, Norm
0.80
1.00
Hablante 1, Norm
1.20
1.40
1.60
Hablante 2, Norm
1.80
11
2.00
I. Acoustic study: Methodology
5.2. Data normalization
F2
1.90
1.40
0.90
0.40
0.40
Media hablantes 1 y 2, Norm
Hablante 1, Norm
0.60
Hablante 2, Norm
0.80
Vocal i Herrero (2010)
1.00
Vocal e Herrero (2010)
1.20
Vocal E Herrero (2010)
1.40
Vocal a Herrero (2010)
Vocal O Herrero (2010)
1.60
Vocal o Herrero (2010)
1.80
Vocal u Herrero (2010)
2.00
12
I. Acoustic study: Methodology
5.2. Data normalization
F2
1.90
1.40
0.90
0.40
0.40
Media hablantes 1 y 2, Norm
0.60
Vocal i Herrero (2010)
0.80
Vocal e Herrero (2010)
1.00
Vocal E Herrero (2010)
1.20
Vocal a Herrero (2010)
1.40
1.60
Vocal O Herrero (2010)
Vocal o Herrero (2010)
1.80
Vocal u Herrero (2010)
2.00
13
Descargar

Rara afis, rara avis